I was raised in the Greek Orthodox church, a part of the family of Eastern Orthodox churches. For the uninitiated, Eastern Orthodoxy is the other branch of Christianity that resulted from the East-West schism. The "Western" church contains Catholicism and all other denominations you are probably familiar with. Unfortunately, throughout my life I have seen how Christianity has gotten a bad rap. And in some sense, that's probably justified. It has been used to justify conquest. Massacres have been committed by warring sects. More recently, there have been numerous child sexual abuse reports and scandals in the Catholic church especially, even one reported as recently as early March of this year, where a multiyear investigation revealed that dozens of priests in Rhode Island molested hundreds of victims over decades. Child sexual abuse happens even in the Eastern Orthodox Church, but it receives much less media attention in the west. A post on the Orthodox Church of America website is remarkably candid about this.

Perhaps the most abundant criticism of Christianity nowadays is that it asks its believers to believe unbelievable things, such as that a man lived who was God and also the Son of God and died by crucifixion and rose from the dead three days later, appearing to his followers. This is a lot to accept, not least because it is entirely incompatible with our modern understanding of reality. I think I speak for many of us, however, in saying that we all still wonder if there is something more beyond our day-to-day rhythm of work, make money, pay bills, watch Netflix, sleep.

Despite the dogma of the church, which I certainly have a difficult time accepting most of the time, I still remain active in the church. This Holy Week, I reflected on why, and the answers boiled down to the following,

1. I have an intuitive recognition of the presence of God in my life.
2. I somehow understand that the teachings of Jesus offer a way to live without anxiety or fear, in a way that is beautiful and selfless. My heart resonates with these teachings as the right way to live, as opposed to trying to hoard as much money as possible to become materially wealthy, and conceited and selfish as a result.
3. I have a deeply rooted community of friends in the church that I grew up with and still stay in touch with. More on this in the second part of this series.

In my view, it is not difficult to experience God. Consider how a tree's branches all move individually when struck by a breeze, and yet the tree dances in the wind. Look around at your greatest friends, and reflect on how they came into your life. I'll bet it was more-or-less by chance, and yet entirely consequential for how you have formed as a person. The warmth of the sun, heating and enlivening all things. The plants and animals that exist for us to marvel at and take care of. When reflecting on such things, I fill up with awe and gratitude, and I hope you do too. But I digress.

In this series of blog posts, I'd like to take you on a tour of spiritually inclined ideas that I find interesting but often don't share with others. In this first post, I'll raise awareness of the error of transactional thinking, and I'll wrap up by trying my hand at writing a parable.

The economy is zero sum

The language of money is universally spoken, but it is not universally expressive. Unpriced externalities, such as environmental side effects of the economic engine, are by definition unpriced--they exist outside of money's lexicon, and yet they materially affect us every day. The cancer burden is projected to keep growing, with harmful chemicals in our supply chain, developed with intent to increase the cheap supply of food, likely playing a significant role. Big Tech profits immensely from harvesting our attention through algorithmically curated infinite-scroll media to serve our brains as many advertisements as humanly possible, but a growing body of research suggests that this has been severely destructive to our attention spans, formation of narrative identity in adolescence, learning outcomes, literacy, and the ability to have spiritual experiences.^{1 2}

As more externalities become apparent with each passing decade--or each passing year at this point--our return on investment from "implementing" the monetary system continues to vanish. The utility of money is that it provides a convenient interface for exchanging value between humans, but in exchange we give up what we actually live for: connection, love, warmth, service. This is not to say that free enterprise is inherently bad; rather, the pursuit of money—through business or otherwise—has blinded us to the needs of our neighbors. Instead of freely giving and freely receiving, a transactional pattern has been imposed onto all things—think "I will only help you if you deserve it or I am compensated."

As Rabbi Michael Knopf points out, one midrash on the Tower of Babel posits that this very same inversion is the error of the people building the Tower of Babel:

... the people were so passionate about building the tower, so convinced were they of its utter necessity, that they paid no mind if a worker on the tower fell to his death; whereas if a brick fell, they would wail and mourn and lament the setback. The peoples’ conviction of the justness of their cause, aided and abetted by the universality of their belief, diminished their humanity and their concern for the welfare of their fellow human beings.

Our economy is the new Tower. And the latest tier of the tower is nearly complete, in the form of artificial intelligence. AI has the potential to manifest immense good, and this is the version of the future I hope to push for and will continue to argue for in my writings. However, if additional tiers of the tower continue to be built, we may invite the wrath of God (perhaps in the form of the dreaded AI takeover…) who wants us to be proud individuals rather than conformers to the pursuit of a single-minded (and absent-minded) goal.

In the New Testament, Jesus also recognizes the perverseness of having money as the fabric of society. A classic example of this can be found in all four canonical gospels: Jesus' entry into the temple and subsequent outrage at those buying and selling within. Here is the version from Mark 11:15-19,

Then they came to Jerusalem. And he entered the temple and began to drive out those who were selling and those who were buying in the temple, and he overturned the tables of the money changers and the seats of those who sold doves, and he would not allow anyone to carry anything through the temple. He was teaching and saying, “Is it not written, ‘My house shall be called a house of prayer for all the nations’? But you have made it a den of robbers.”

This is one of the very few times in the New Testament where Jesus gets physically aggressive. It's a powerful image, and a relevant one. Our world is the new den and many corporations are the new robbers. Exploitation in the name of money happens every day billions of times over on Earth today. In America, our clothes are bought with modern slave labor.. Our chocolate is bought with child labor in Africa.

Money can no longer be the interface if we are to live meaningful, moral lives. A life lived for money is zero sum. In death, money disappears, and for every dollar extracted from the economy, there is a potential friend that could have been made, or an idea that could have been shared, or a person in need who could have been helped.

Against the zero-sum life

This ethos of rejection of money as a mediator of value is illustrated further in how the early followers, the first people who can be called Christians, conducted themselves in the book of Acts. I have always found the early chapters of Acts to be particularly heartwarming, because they describe humans living in communion with each other and their surroundings. For example,³

They devoted themselves to the apostles’ teaching and to fellowship, to the breaking of bread and to prayer. Everyone was filled with awe at the many wonders and signs performed by the apostles. All the believers were together and had everything in common. They sold property and possessions to give to anyone who had need. Every day they continued to meet together in the temple courts. They broke bread in their homes and ate together with glad and sincere hearts, praising God and enjoying the favor of all the people. And the Lord added to their number daily those who were being saved.

This is diametrically opposed to the norms of our world. Taking this depiction at face value, it is clear that ownership as we conceive of it now did not exist in the early Christian community.⁴ The disciples' way of life sent ripples out into the world. Their commitment to nontransactional thinking allowed them to create a movement larger than themselves, so much so that I'm still writing about it now, 2000 years later.

Those who don't think transactionally have a willingness to be generous, to love, and to be unconcerned with what they will receive in return. This inspires everyone around them (I hope) to be better. If you aren't religious, now perhaps you can understand why people choose to go to church on Sunday to hear how Jesus and his followers interacted with the world.

The good news is that developing nontransactionality in thought is simple. Like anything else, it is a skill that can be practiced. Freely give, and freely receive.⁵ When someone is in need, do something for them without worrying about whether they will do something for you. Don't second guess anything, so that you do not "let your left hand know what your right hand is doing."⁶ When someone helps you, accept it with gratitude and don't worry about "repaying them" as if you owe them a debt, unless they ask you to.

This might seem like a losing proposition: we get taken advantage of, we give more than we take, how does this profit us? This is transactional thought programming at work. Why do we believe we get taken advantage of when we give more than we take? What "profit" are we looking for? There isn't a logical answer to this question, despite the desire I have to articulate it clearly for you on this page. Ultimately, letting go of transactionality is a choice.

I'm advocating for this simply because I believe our world is better for it. And you'll feel better for it. That is my platform.

Sendoff

Go forth, and be active rather than passive in the world. This is one of the themes of the "different way" that I plan on building up in this series of posts. Letting go of transactionality means becoming an active participant in the lives of other people, rather than passively allowing your trajectory to be guided by the transactional mode of seeking profit and paying back debts.

I hope to respond to my own call as well. I know I still have a journey ahead of me toward living actively and nontransactionally. This blog is part of that effort.

I'm sure this won't land for everyone, and that's okay. But if it lands for you, drop me an email and subscribe to my newsletter with the box at the bottom of the page. One of my goals with this blog is to start a community around this sort of thinking.

In the next part, I will recount a particularly moving experience from my adolescence which revealed to me the boundless love that can exist in social bonds between humans. Then, I will reflect on how modern Christianity may be turning into "just in case" religion, and propose an alternative path forward.

A parable

On a street there are three houses, each with a man living in it.

The first house has an exterior that is unkempt. The roof is in need of repair. But the poor and needy in the town know that this man will invite them in, so they come to his house, and he invited them in and feeds them and clothes them and gives them whatever they ask.

The second house is well kept externally. Visitors seldom enter the house and the man inside seldom ventures outside. This man has friends from across the city that he occasionally invites over, and they invite him over from time to time as well.

The third house is also well kept, and this man is always outside, inviting strangers passing by to come in. He even walks over to the doors of the first and second houses to invite those men to share a meal with him. “We are neighbors,” he says, “why should we not be friends?”

A curious traveler came to this city and stopped at each of the three houses. In the first house, he was received with kindness because he knocked on the door. The man inside gave the traveler bread and some money. The second house did not open its door to the traveler even after persistent knocking. The man in the third house ran out to greet the traveler as he was walking towards the house. The man then invited the traveler inside where he already had others from around the city at the table. They laughed and ate and shared themselves with each other.

The traveler asked the man, “why do you do this?”

The man replied, “I give food and drink and whatever I can offer so that thieves do not steal. I offer up my house so that vagrants have a place to rest their heads. I regard all as friends so that no one is lonely. I keep my house so that it is inviting and run out to greet people so that no one is afraid to ask anything of me.”

The traveler left the third house full and content and filled with happiness. He said, “I will never forget the man in the third house.”

I am on day four of no coffee after drinking it almost daily for several years. Sunday was rough. I had a pressure headache all day and was in a bad mood. Thinking felt hard. Monday was better, but still had a headache. I caved and had a black tea around lunchtime so that I could get my work done. On Tuesday, it seemed that my brain was relearning how to clear out adenosine, or downregulating adenosine receptors, or whatever the biological mechanism looks like that makes me feel more able to stay awake and direct my mental energy toward something. Today is Wednesday. I still have a slight headache but overall I feel fine.

Interestingly, I've found it easier to stay up later and get up earlier without caffeine, because I avoid the crash late at night when the caffeine leaves my system, and I don't feel sluggish in the morning before I have my morning coffee. It feels like a win all around.

I also gave up alcohol for Lent (not entirely; I've occasionally been drinking a glass of beer or wine but very limited in quantity, and no hard liquor). It has made me realize how even a little bit of alcohol too late in the evening has a deleterious effect on my sleep.

Overall, I like this mental state of not having peaks and troughs in my consciousness. Caffeine may make me feel able to do more in the short term, but as my withdrawals seem to show, unearned gains in the present always borrow from the future, and that debt has to be repaid, sometimes with interest. Same story with alcohol. On weekends when I would go out with friends and drink for multiple nights in a row, I would show up to work on Monday having lost the thread I was tracing the previous Friday. This doesn't happen as much with no alcohol. I suspect its effects on sleep and memory may have something to do with this.

I think I'll try to keep this up for as long as I can. I feel more energetic and more creative. Of course, no shame if you enjoy caffeine and/or alcohol (in moderation). This is just what works for me.

Enjoy your day, dear reader!

I work in software, and since the beginning of this year, 2026, I have been investing in deeply integrating AI into my workflow. AI has gotten strikingly good at coding, to the point where it writes more than 90% of my code, and many of my coworkers at Stripe are reporting a similar experience. Everyone in software that I talk to agrees: Anthropic's Opus 4.6 unleashed a vibe shift in coding capability. A world where software engineering is mostly automated feels within reach.

Although there are many questions unanswered about the safety of mass AI adoption, I think there is great potential for the replacement of jobs by AI to create a remarkably improved stan1d of living, where humans can be free of most material constraints and return to a state of creativity and community-oriented living. The transitional period might be painful, but I am generally optimistic about what a fully AI-driven future could look like. We can maximize our chances of success if we pair a push for powerful AI-driven job automation with Scalable Formal Oversight (SFO).

If we don't act now, this generational opportunity might fade away. One way it could fade is if inference costs rise substantially. Recently, I caught wind of news that a large software company (name omitted for privacy) is pumping the brakes on AI spend. The team responsible for rolling out AI tools must have received a directive to start constraining costs because the company's bottom line was starting to feel it, and they forced all engineers to use a lower tier model unknowingly. In response, several engineers at the company grumbled about how the lower model tier's output was noticeably worse, and it led to wasted cycles spent on reviewing and correcting the worse model's output.

This got me thinking about whether the AI adoption and job replacement scenarios painted in pieces like the Citrini report and Matt Schumer's now infamous Something Big Is Happening post are realistic in our current world.

My take is that AI labor replacement is not going to happen yet. AI inference will get significantly more expensive due to macro and geopolitical risks and the fleeting nature of investor subsidies of inference. If that happens, the demand for inference that's fueling the AI buildout could evaporate due to prohibitively high costs, thus completing the boom-bust cycle. I'm disappointed about this because I believe that this technology has the potential to transform the way we work and live, but all hope is not lost; there are several strategic directions in which we can collectively push to maximize the likelihood of a good outcome.

By the end of this post, I hope to reinvigorate in you (and myself, quite frankly) a revolutionary spirit with some thoughts about what actions we can take to try to generate escape velocity from the boring future where AI becomes a half-decent tool that makes work more annoying and exhausting. But to start, here are the main pillars of my argument:

1. AI buildout is bottlenecked: Constraints on the manufacturing of critical inputs limit the AI buildout. Recent geopolitical instability and its effects on energy prices threaten to further dampen progress.

2. Inference costs are being held artificially low: Investor money is subsidizing the free tiers of frontier labs. We should expect inference costs to rise substantially if the frontier labs providing the models we know and love want to stay in business.

3. Humans are now the bottleneck more than ever: Investment in AI tools are reportedly not generating ROI for many enterprises. AI helps workers do more, but the limits of cognitive load of managing and communicating with AI agents in parallel are becoming apparent.

Expect inference costs to increase

When an environment changes, systems operating in that environment either fail or adapt. The release of ChatGPT in 2022 changed the environment, and three and a half years later, our economic system is still in the process of adapting.

How quickly different parts of the economy can adapt is variable. Software adapts very quickly, hardware, more slowly, and physical manufacturing infrastructure adapts at a snail's pace. It is therefore unsurprising that the chokepoints of the AI buildout are in the realm of the physical: power and silicon.

Frontier labs are also underpricing inference, drawing down their prodigious cash reserves to cover their costs. If the macro risk flares up, then capital might dry up before profitability is reached.

Risk 1: Macro

If you're tapped into the sphere of popular tech podcasts, such as the Cheeky Pint podcast by John Collison, cofounder of Stripe, you may have heard Elon Musk in a recent episode talking about a severe shortage of the turbines that are used to convert natural gas into electricity. Gas-fired turbine lead times have soared to as much as seven years. That means that if you aren't already in the queue, you're waiting close to a decade to getting just the turbines, which are necessary but not sufficient to generate electricity from natural gas.

The supply of natural gas itself, along with oil, are also being threatened. Recent US military action in Iran has effectively closed the strait of Hormuz, which is disrupting global oil and liquefied natural gas (LNG) supply. According to Reuters, the situation in the strait has removed around 400 million barrels from the market, "triggering price increases of around 50%." This crisis may be transient, and we will see how things play out in the coming weeks, but it highlights the US and world economies' sensitivity to global conflict.

These circumstances may restrict how quickly new data centers can come online, depending on what fuel source they choose. Gartner estimates 2025 data center power consumption 448 TWh, and they estimate it to grow to 980 TWh in 2030. In their model, power consumption by data centers more than doubles in five years, while growth in power generation across the board in the US is alarmingly stagnant.¹

As for silicon, TSMC, a Taiwanese corporation, represents a single point of failure for the US tech industry. In Q2 2025, TSMC captured a record 70.2% of chip foundry market share. According to Nasdaq, TSMC has an even higher share, 90%, of advanced chip manufacturing, which includes 3-nanometer chips that are becoming standard. TSMC poses obvious geopolitical risk as well, as is a mere 100 miles off the coast of China. One well targeted and well timed attack could disrupt AI adoption by limiting how much compute can get built out in the short term. There is no telling how long it would take to reach original production capacity again if there were significant damage to the dense network of fabs, raw materials suppliers, and research institutions on Taiwan's west coast.

Ostensibly in response to these risks, the US government pressured TSMC to spin up a fab in Arizona. The fab has been built, but TSMC Arizona already lags behind Taiwan in the set of manufacturing processes that are available. TSMC is currently capable of lithography down to 2 nanometer scale, and cloud service providers are rushing to adopt it, but 2nm and the older 3nm processes are slated to be made available in the Arizona plant in 2029 and 2027, respectively. In Taiwan, a 1nm fab is already set for Tainan. US-based fabs have a lot of catch-up to do. We must have parity with the state of the art in chip manufacturing before the supply chain and geographical dependencies can be eliminated.

These are far from the only risks to chip fabrication either. Just as I was drafting this post, a notification lit up my phone: "An Invisible Bottleneck: A Helium Shortage Threatens the Chip Industry".

It becomes hard to justify underpricing inference when the cost of the physical utility infrastructure might blow up, and it's impossible to justify underpricing when there is no capital to subsidize inference. This brings us to the next risk: capital.

Risk 2: Capital well runs dry

Let's look at some numbers, using OpenAI as an example. Several different sources cite different revenue and cost figures for OpenAI, and it is a private company, so it is difficult to know their true financials with certainty. At a minimum, however, we can certainly say OpenAI is losing money at an astounding clip. Epoch AI estimates that OpenAI was significantly unprofitable in 2024, making $3.7 billion revenue on $1.8 billion inference spend and roughly $6 billion total compute spend, including inference. This implies a plausible 50% gross margin on inference, but a negative net margin that investors are covering.

In 2025, OpenAI reported $4.3 billion in 2025H1 and projected $13 billion for full year. "Against that revenue, they’re spending approximately $22 billion," representing a shortfall of $9 billion that investors are subsidizing. This shortfall is set to increase, as further research from Epoch AI shows that on a large scale, training compute costs have been doubling every 8 months, or 2.7x per year.

OpenAI and their investors are well aware of the capital requirements to keep the lights on with inference as well as in the R&D department. The company is fresh off a record fundraise of $120 billion. Even with that much capital, OpenAI still has to be judicious. Just last week, OpenAI decided to shut down Sora, the video generation app released several months ago. OpenAI's CFO, Sarah Friar, said on CNBC, "We are just facing a lack of compute." TechCrunch cites the Wall Street Journal's analysis of the situation, stating that "Claude Code, in particular, was eating OpenAI’s lunch."

If the macro risks increase, leading to costlier compute, OpenAI may not be able to make the rounds with investors to raise eyewatering amounts of money much longer. Furthermore, all this is moot if these models don't perform in real-world deployments. With 61% of senior business leaders feeling increased pressure to prove ROI compared to a year ago, and 71% of global CIOs saying their AI budgets would be cut if value from AI couldn't be demonstrated within 2 years, fiscal reality may start settling in soon. If poor performance metrics coincide with a compute cost spike, AI investment and adoption fervor could rapidly deteriorate.

The competition is fierce. I predict that being able to manage costs will soon become a moat. A good example of this is Cursor. They leveraged their unique dataset of software engineer conversations with AI to build in-house models that are domain-specific for coding and far less expensive to run. If capital runs out, labs like OpenAI and Anthropic will have to either cut costs or significantly increase the cost of using their products, restricting the accessibility of powerful AI. If only mediocre AI remains accessible, we will settle into a stationary point of productive capacity where humans remain the bottleneck.

Humans are more of a bottleneck than ever

Working with AI tools has a strange allure, almost an addictive quality. The gap between dream and reality has shortened, so the dopamine hit we get from completing a task is closer and more predictably reachable.

This thrill has enabled us to take on more work. I can attest that there is a thrill to being able to kick off a new agent session when an idea comes to my head instead of having to invest the mental effort of thinking through what I'm trying to do. This is a pattern that is appearing across the economy. According to research published in HBR,

On their own initiative workers did more because AI made "doing more" feel possible, accessible, and in many cases intrinsically rewarding.

The researchers warn, however, that this early period of experimentation could morph into cognitive overload, reducing quality of work output and decision-making in the long term.

Humans have a fundamentally serial input-output interface. We fail at multitasking, and there is a well-established body of research that corroborates this.² AI, on the other hand, can be virtually infinitely parallelized, producing more output per unit time than is humanly possible to review and understand in a timely fashion, and trying to switch between parallel threads deleteriously imposes cognitive overhead. This effect, sometimes referred to as "AI Brain Fry" is analogous to thrashing on computers. It is a slowdown caused by the serialization of infinite parallel AI workstreams, and the temptation to add more workstreams to "be more productive."

This all points to the need to develop a better working relationship with AI. Perhaps we need to be able to entrust it with longer horizon tasks so that we aren't as crushed by the mathematics of how much effort it takes for our brains to switch context and enter deep focus mode. I'll take this one step further: task time horizon where AI can reliably (>>90%) execute tasks independently is the most important dimension for job replacement. It is more important than cost because capital can be acquired if the return-on-investment potential is there (i.e. loans).

Reliable long horizon task completion, however, still eludes even the most powerful models. According to the METR task completion time horizon evaluation, Opus 4.6 can complete a task that would take a human 1 hour and 10 minutes with only 80% reliability.³ 80% reliability is significantly less reliable than a well-calibrated human in a workplace. Correcting the output of AI 20% of the time for long horizon tasks and having to context switch frequently to manage many short horizon tasks doesn't sound like a healthy, productive working model to me.

So for the time being, it seems that we are stuck in this rut where AI works decently well but leaves us cognitively exhausted due to managing it. If future AI can achieve reliability figures in the high 90s but only with extremely high cost, such that ROI still remains questionable, then the fantasies of Citrini-style job replacement vanish.

How can we manifest the "good ending?"

Now, as promised, I would like to paint a different picture and suggest strategic directions in which we can collectively push to manifest what I'm calling the "good ending," where job replacement does happen and it enables us to become our true creative, soulful, human selves.

Upon reading the above, you might be wondering, "why is the current state suboptimal?" and "how does job replacement enable this vision of yours?" These are fair questions, and it is true that these are not guarantees. Work gives many people a sense of purpose, and many people may not want to lose that. However, community involvement can also grant a sense of purpose. I have felt this in my own life. Spending time with my family and community and helping them out motivates me like few other things in the world. I think one of the tragedies of our modern economic system is that community-building has fallen by the wayside.

Alongside community is creativity which seems uniquely able to form within us a moral conscience. For exaxmple, Uncle Tom’s Cabin by Harriet Beecher Stowe sent waves through the northern states, greatly furthering the abolitionist cause. Literature, art, and other creative works transmit memories of human pain and suffering of the past, helping us learn and reflect in the process. I believe humanity benefits from being able to spend lots of time creating and reflecting on creative works.

AI displacement of jobs enables community-orientation and creativity simply by returning to us the choice of what we do with our time. Many of us may choose to keep working, and that's fine. Many jobs have space for the creative work I described.⁴ Many of us may choose other pursuits. It is not a given that our governmental systems will respond in kind by allocating resources appropriately in a post-work world, but I remain optimistic that we can prosper nonetheless.⁵ Without needing to sacrifice community to make a living, we can turn towards our neighbors and connect with them. We can generate new cultures and traditions.

This line of thinking calls to mind the Greek culture that I grew up in. There are rich regional traditions of music, dance, cuisine, customary celebrations, and other cultural artifacts generated from generations of playful experimentation.

These thoughts have parallels with and draw inspiration from this post by Jordan Hall on X. Shoutout to my friend Nick Makiej for finding it and sharing it with me. Now, the following are the directions I suggest we push toward to generate this reality.

1. Create

New generations of AI models are currently getting trained on immense amounts of AI slop content that was generated by past generations of models. Epoch AI predicts we will soon run out of human-generated data to train on. Andrej Karpathy expressed concern about this as well in his discussion of the notion of "entropy" in his interview with Dwarkesh Patel. Entropy is a formal term in information theory, but Andrej was using it in more of informal sense in the interview, to denote the quality of "newness" of information. Trainers of models have been exploring using LLMs to generate synthetic data to train on. Karpathy is skeptical of this approach, however.

Here is an excerpt from the interview:

The LLMs, when they come off, they’re what we call “collapsed.” They have a collapsed data distribution. One easy way to see it is to go to ChatGPT and ask it, “Tell me a joke.” It only has like three jokes. It’s not giving you the whole breadth of possible jokes. It knows like three jokes. They’re silently collapsed.

You’re not getting the richness and the diversity and the entropy from these models as you would get from humans. Humans are a lot noisier, but at least they’re not biased, in a statistical sense. They’re not silently collapsed. They maintain a huge amount of entropy. So how do you get synthetic data generation to work despite the collapse and while maintaining the entropy? That’s a research problem.

I tried this myself to see if Karpathy was right, using the free version of ChatGPT on March 22nd, 2026 (this isn't a very reproducible experiment, because OpenAI hides which model is being used on the free version). On the fourth try, ChatGPT repeated the same joke as the third try:

The upshot here is that we as humans still have a lot to offer. Maybe inference doesn't have to decrease substantially in cost if we can collectively provide current models with more entropy per token.

Each of us contains a deep well of entropy, informed by our unique personal experiences and story. This is what inspired me to start this blog. I used AI to research this piece, but I wrote every word on this page by hand because I wanted this blog to reflect my entropic voice.

Go forth, create, and create with high quality.

2. Participate in research to make AI more cost effective to train and run

The further we drive down the cost of inference, the more likely it is that we'll be able to generate escape velocity and drive meaningful adoption of AI in workplaces and among people at the edge.

Research in quantization, distillation, and pruning has made strides toward making LLM inference runnable on cheap hardware such as phones, but with a slight accuracy penalty. Figuring out how to mitigate that tradeoff is an interesting research direction.

I think alternatives to the LLM architecture are also underexplored. Drawing from the Mamba architecture, for instance, could lead to significant efficiency improvements, as Mamba's attention has linear time complexity whereas standard transformer model attention layers have quadratic time complexity.

My good friend Dana recently joined Subconscious, a startup leveraging research out of MIT which proposes a codesigned thread inference architecture and runtime for long horizon tasks. Codesign, though harder to implement and not generalizable, can have tremendous efficiency benefits because each part of the system is aware of how it needs to interact with the others to maximize efficiency and accuracy. Architecture-runtime codesign is a great first step. I would love to see this taken a step further toward runtime-chip codesign as well. MatX is a startup building chips for inference, and I expect they've done some thinking about this.

A nice consequence of all this is that cheap inference that can be run with consumer hardware at the edge also solves the energy bottleneck. The power grid is designed to handle marginal load increases at its edges, but not well equipped to handle a massive spike in load at a single geographical point. This might even resolve some of the silicon bottleneck too; consider how much compute there is around us lying dormant on devices like smartphones, tablets, even our laptops.

There is lots of interesting work to be done here, and I look forward to seeing it.

3. Participate in the Scalable Formal Oversight initiative

The obvious risk posed when racing towards the "good ending" is the AI doomer's fantasy: what if AI takes over and starts misbehaving?

This is not a fantasy. The AI of today misbehaves often. Large language models are known to engage in deceptive behavior to make human evaluators think that they are more aligned with human intentions than they actually are. Code that agents output can be unsafe and destructive. A verification harness for the actions that AI agents take is crucial. In my view, it is a prerequisite for any significant job replacement in safety critical domains, and it is likely necessary in order to actually enable large scale replacement in software engineering, for instance, where high level goals can be very long horizon, and contexts can grow very large. Agents are likely prone to mistakes that may be avoidable with verification guardrails.

Read Max von Hippel's post Scalable Formal Oversight, and take it seriously!

4. Build community

At the end of the day, community is all we have. If AI frees us from most material constraints, then we can return to a world of playful creativity, cheer, and communal bonds. This is a world I'd love to live in.

If this resonates with you, get in touch!

Why I could be wrong

To recapitulate: my core thesis is that we can expect AI inference to get significantly more expensive, due to macro risks and venture capital inference subsidies running out. Cheap access to AI inference at the frontier of capability is necessary to support nontrivial levels of job replacement, so we will likely land in a situation where fervor around AI will slow greatly.

Here are some rapid-fire reasons why I could be wrong:

1. Perhaps progress in making inference cheaper at the hardware level will speed up. MatX is a chip startup hoping to reinvent how AI inference is done at a hardware level, which is compelling, and perhaps they will have some breakthrough.
2. Perhaps the geopolitical risks of today will evaporate tomorrow.
3. Perhaps the buildout will proceed more smoothly than I think it will, with TSMC's Arizona buildout proceeding as planned or better than planned.
4. Perhaps the problem isn't inference cost, but context management or other algorithmic inefficiencies (another shoutout to Dana at Subconscious), and better context management is an active research area. We could generate escape velocity if we can solve these challenges and unlock more consistent ROI.
5. Perhaps most industries won't actually require highly capable LLMs to be mostly automated.
6. Perhaps people are still learning how to most effectively use AI. A recent report from Anthropic suggests this very effect. Longer tenure users of Claude are 4% more likely to have successful conversations, even after controlling for model type, use case, and country/language. Multitasking is also a skill that appears to be trainable.

There are probably more reasons. Send me an email if you have a comment! This post is an essay⁶ in the true sense of the word.

Each of these probably deserves more exploration, but I did not have the time or space to get them into this piece. I may return to some of these pushbacks in future posts, so stay tuned!

Final word

Thank you for making it this far. This is the first blog post that I'm officially publishing, so I would appreciate any and all feedback. Please drop me an email at s.xifaras999@gmail.com if you would like to leave feedback, positive or critical, or just want to chat. If you're in the northeastern US (NYC-Boston range), let me know and I'd love to explore an in-person meeting!

Victor and Nick, thank you for your helpful feedback on drafts of this post.

I've been meaning to write more for a while. Up to this point, this site has been a static page — a CV with links — but I want a place to think through ideas in public.

My goal is to write about whatever comes to me and piques my interest. I expect that my first few posts will be a bit rough, and that's okay. I am being intentional about cultivating high follow-through; I am not letting perfect be the enemy of good.

What to Expect

Posts will be irregular. I'm not committing to a schedule, just to writing when there's something worth saying.

If you have thoughts on anything I post, or just want to say hi, please reach out via email.